Airbnb - Boston

# Import Neccessary libraries
import pandas as pd 
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sb
import math
import time


# Import library for VIF
from statsmodels.stats.outliers_influence import variance_inflation_factor
from sklearn.linear_model import ElasticNet
from sklearn import preprocessing
from sklearn.metrics import *


import folium
from folium import plugins
from folium.plugins import HeatMap

ur = "http://data.insideairbnb.com/united-states/ma/boston/2021-12-17/data/listings.csv.gz"
ur1 = "http://data.insideairbnb.com/united-states/ma/boston/2021-12-17/visualisations/listings.csv"

We have imported pandas ,numpy for basic data analysis. Seaborn and matplotlib for data visualization. Folium for maps.

# Loading Dataset
boston = pd.read_csv(ur,compression='gzip',low_memory=False)
boston2 = pd.read_csv(ur1)

Data Exploration , Visualization and Processing

### Created a Subset of data
dat = boston[['host_id','accommodates','bedrooms','beds','amenities']].copy()
boston3 = pd.merge(boston2,dat)
boston3.tail()

	id	name	host_id	host_name	neighbourhood_group	neighbourhood	latitude	longitude	room_type	price	...	last_review	reviews_per_month	calculated_host_listings_count	availability_365	number_of_reviews_ltm	license	accommodates	bedrooms	beds	amenities
107384	53772114	Entire Modern Apartment in Downtown Boston 2 R...	434990435	Luke	NaN	South End	42.344829	-71.062141	Entire home/apt	139	...	NaN	NaN	5	89	0	str-000000	2	1.0	1.0	["Smoke alarm", "Air conditioning", "Pool", "W...
107385	53772114	Entire Modern Apartment in Downtown Boston 2 R...	434990435	Luke	NaN	South End	42.344829	-71.062141	Entire home/apt	139	...	NaN	NaN	5	89	0	str-000000	2	1.0	1.0	["Smoke alarm", "Air conditioning", "Pool", "W...
107386	53772114	Entire Modern Apartment in Downtown Boston 2 R...	434990435	Luke	NaN	South End	42.344829	-71.062141	Entire home/apt	139	...	NaN	NaN	5	89	0	str-000000	4	1.0	1.0	["Paid parking on premises", "TV", "Outdoor di...
107387	53772114	Entire Modern Apartment in Downtown Boston 2 R...	434990435	Luke	NaN	South End	42.344829	-71.062141	Entire home/apt	139	...	NaN	NaN	5	89	0	str-000000	6	1.0	1.0	["Outdoor dining area", "Smoke alarm", "Air co...
107388	53756737	The Arbor Retreat	42715907	Emma	NaN	Jamaica Plain	42.304864	-71.129771	Entire home/apt	112	...	NaN	NaN	1	231	0	STR-463838	2	1.0	1.0	["Dishes and silverware", "Cooking basics", "R...

5 rows × 22 columns

### Calculating Summary Statistics
boston3.describe()

	id	host_id	neighbourhood_group	latitude	longitude	price	minimum_nights	number_of_reviews	reviews_per_month	calculated_host_listings_count	availability_365	number_of_reviews_ltm	accommodates	bedrooms	beds
count	1.073890e+05	1.073890e+05	0.0	107389.000000	107389.000000	107389.00000	107389.000000	107389.000000	29052.000000	107389.000000	107389.000000	107389.000000	107389.000000	88219.000000	82280.000000
mean	4.461479e+07	1.758024e+08	NaN	42.342998	-71.084340	147.84707	51.843345	11.181425	1.655152	127.744713	275.678971	3.312136	2.281034	1.246920	1.370017
std	1.192542e+07	1.221205e+08	NaN	0.019351	0.033584	116.54091	36.352791	38.862501	1.847476	71.816091	105.919076	11.022718	1.483380	0.586296	0.845982
min	3.781000e+03	4.804000e+03	NaN	42.235330	-71.172520	0.00000	1.000000	0.000000	0.010000	1.000000	0.000000	0.000000	0.000000	1.000000	1.000000
25%	4.208014e+07	1.074344e+08	NaN	42.335150	-71.101850	50.00000	30.000000	0.000000	0.220000	46.000000	226.000000	0.000000	1.000000	1.000000	1.000000
50%	4.961078e+07	1.074344e+08	NaN	42.347970	-71.070840	149.00000	32.000000	0.000000	1.000000	168.000000	318.000000	0.000000	2.000000	1.000000	1.000000
75%	5.227869e+07	2.978601e+08	NaN	42.355180	-71.061640	212.00000	91.000000	1.000000	2.410000	194.000000	364.000000	0.000000	3.000000	1.000000	2.000000
max	5.383997e+07	4.349904e+08	NaN	42.392790	-70.997810	10000.00000	1000.000000	678.000000	27.170000	194.000000	365.000000	298.000000	16.000000	13.000000	22.000000

### Precise Summary of the Dataframe 
boston3.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 107389 entries, 0 to 107388
Data columns (total 22 columns):
 #   Column                          Non-Null Count   Dtype  
---  ------                          --------------   -----  
 0   id                              107389 non-null  int64  
 1   name                            107389 non-null  object 
 2   host_id                         107389 non-null  int64  
 3   host_name                       61820 non-null   object 
 4   neighbourhood_group             0 non-null       float64
 5   neighbourhood                   107389 non-null  object 
 6   latitude                        107389 non-null  float64
 7   longitude                       107389 non-null  float64
 8   room_type                       107389 non-null  object 
 9   price                           107389 non-null  int64  
 10  minimum_nights                  107389 non-null  int64  
 11  number_of_reviews               107389 non-null  int64  
 12  last_review                     29052 non-null   object 
 13  reviews_per_month               29052 non-null   float64
 14  calculated_host_listings_count  107389 non-null  int64  
 15  availability_365                107389 non-null  int64  
 16  number_of_reviews_ltm           107389 non-null  int64  
 17  license                         23450 non-null   object 
 18  accommodates                    107389 non-null  int64  
 19  bedrooms                        88219 non-null   float64
 20  beds                            82280 non-null   float64
 21  amenities                       107389 non-null  object 
dtypes: float64(6), int64(9), object(7)
memory usage: 18.8+ MB

### Check for null values
boston3.isna().sum()

id                                     0
name                                   0
host_id                                0
host_name                          45569
neighbourhood_group               107389
neighbourhood                          0
latitude                               0
longitude                              0
room_type                              0
price                                  0
minimum_nights                         0
number_of_reviews                      0
last_review                        78337
reviews_per_month                  78337
calculated_host_listings_count         0
availability_365                       0
number_of_reviews_ltm                  0
license                            83939
accommodates                           0
bedrooms                           19170
beds                               25109
amenities                              0
dtype: int64

### Percentage of null rows in each column 
def missing(datas):
    print (round((datas.isnull().sum() * 100/ len(datas)),2).sort_values(ascending=False))

missing(boston3)

neighbourhood_group               100.00
license                            78.16
last_review                        72.95
reviews_per_month                  72.95
host_name                          42.43
beds                               23.38
bedrooms                           17.85
id                                  0.00
accommodates                        0.00
number_of_reviews_ltm               0.00
availability_365                    0.00
calculated_host_listings_count      0.00
number_of_reviews                   0.00
name                                0.00
minimum_nights                      0.00
price                               0.00
room_type                           0.00
longitude                           0.00
latitude                            0.00
neighbourhood                       0.00
host_id                             0.00
amenities                           0.00
dtype: float64

boston3.replace({'f': 0, 't': 1},inplace = True) 
boston3.hist(bins=50, figsize=(10,10))
plt.savefig('distribution.png', dpi=650, bbox_inches='tight')
plt.show()

We can drop the columns with less categories.Checking whether boolean and categorical features contain sufficient numbers of instances in each category to make them worth including.It can be seen that several columns only contain one category and can be dropped while preprocessing.

Data Preprocessing

# To protect the privacy of the hosts and reviewers, we drop them as well as the unwanted coulumns
boston3.drop(['id','host_name','last_review','license',], axis=1, inplace=True)

boston3.drop(['neighbourhood_group'], axis=1, inplace=True)

boston3.isnull().sum()

name                                  0
host_id                               0
neighbourhood                         0
latitude                              0
longitude                             0
room_type                             0
price                                 0
minimum_nights                        0
number_of_reviews                     0
reviews_per_month                 78337
calculated_host_listings_count        0
availability_365                      0
number_of_reviews_ltm                 0
accommodates                          0
bedrooms                          19170
beds                              25109
amenities                             0
dtype: int64

#Percentage of null rows in each column
def missing(datas):
    print (round((datas.isnull().sum() * 100/ len(datas)),2).sort_values(ascending=False))

missing(boston3)

reviews_per_month                 72.95
beds                              23.38
bedrooms                          17.85
name                               0.00
accommodates                       0.00
number_of_reviews_ltm              0.00
availability_365                   0.00
calculated_host_listings_count     0.00
number_of_reviews                  0.00
host_id                            0.00
minimum_nights                     0.00
price                              0.00
room_type                          0.00
longitude                          0.00
latitude                           0.00
neighbourhood                      0.00
amenities                          0.00
dtype: float64

### Filling the null values with mean and median

boston3['reviews_per_month'].fillna(boston3['reviews_per_month'].mean(), inplace=True)
boston3['beds'].fillna(boston3['beds'].median(), inplace=True)
boston3['bedrooms'].fillna(boston3['bedrooms'].median(), inplace=True)

missing(boston3)

name                              0.0
reviews_per_month                 0.0
beds                              0.0
bedrooms                          0.0
accommodates                      0.0
number_of_reviews_ltm             0.0
availability_365                  0.0
calculated_host_listings_count    0.0
number_of_reviews                 0.0
host_id                           0.0
minimum_nights                    0.0
price                             0.0
room_type                         0.0
longitude                         0.0
latitude                          0.0
neighbourhood                     0.0
amenities                         0.0
dtype: float64

bosub = boston3[["price","number_of_reviews","reviews_per_month","calculated_host_listings_count"]]

### Plot to check the outliers
boston1=bosub.select_dtypes(exclude=['object'])

for column in boston1:
        plt.figure(figsize=(10,2))
        sb.boxplot(data=boston1, x=column)

#Since our target variable is the price, we have standardized the values such that the mean is zero and the values 
# are in one standarad deviation
x_data = boston3.price
standard = preprocessing.scale(x_data)
print(standard)

[-0.19604426 -0.41914292 -0.41914292 ... -0.07591421 -0.07591421
 -0.30759359]

boston3.amenities[:1].values

array(['["Cable TV", "Dishes and silverware", "Cooking basics", "Patio or balcony", "Hair dryer", "Refrigerator", "Hot water", "Dedicated workspace", "Bed linens", "Wifi", "Essentials", "Washer", "Air conditioning", "TV with standard cable", "Oven", "Stove", "Dryer", "Smoke alarm", "Extra pillows and blankets", "Hangers", "Carbon monoxide alarm", "Dishwasher", "Heating", "Shampoo", "Free parking on premises", "Iron", "Kitchen", "Coffee maker", "Microwave", "Long term stays allowed", "Free street parking"]'],
      dtype=object)

amenities_list = list(boston3.amenities)
amenities_list_string = " ".join(amenities_list)
amenities_list_string = amenities_list_string.replace('{', '')
amenities_list_string = amenities_list_string.replace('}', ',')
amenities_list_string = amenities_list_string.replace('[', '')
amenities_list_string = amenities_list_string.replace(']', ',')
amenities_list_string = amenities_list_string.replace('"', '')
amenities_set = [x.strip() for x in amenities_list_string.split(',')]
amenities_set = set(amenities_set)
amenities_set

{'',
 '2-5 years old',
 '21\\ HDTV with Roku',
 '24-hour fitness center',
 '32\\ HDTV',
 '32\\ HDTV with',
 '32\\ HDTV with Amazon Prime Video',
 '32\\ HDTV with Netflix',
 '32\\ HDTV with standard cable',
 '36\\ HDTV with Apple TV',
 '37\\ HDTV with Amazon Prime Video',
 '40\\ HDTV with Amazon Prime Video',
 '40\\ HDTV with Apple TV',
 '40\\ HDTV with Roku',
 '40\\ HDTV with standard cable',
 '40\\ TV',
 '40\\ TV with Apple TV',
 '42\\ HDTV',
 '42\\ HDTV with Amazon Prime Video',
 '42\\ HDTV with Roku',
 '42\\ HDTV with standard cable',
 '43\\ HDTV with Amazon Prime Video',
 '43\\ HDTV with Roku',
 '45\\ HDTV with premium cable',
 '46\\ HDTV with Amazon Prime Video',
 '48\\ HDTV',
 '48\\ HDTV with',
 '5-10 years old',
 '50\\ HDTV with',
 '50\\ HDTV with Chromecast',
 '50\\ HDTV with Roku',
 '50\\ HDTV with premium cable',
 '50\\ HDTV with standard cable',
 '52\\ HDTV',
 '52\\ HDTV with standard cable',
 '52\\ TV with standard cable',
 '54\\ HDTV with Netflix',
 '54\\ HDTV with standard cable',
 '55\\ HDTV',
 '55\\ HDTV with Amazon Prime Video',
 '55\\ HDTV with Apple TV',
 '55\\ HDTV with HBO Max',
 '55\\ HDTV with Netflix',
 '55\\ HDTV with Roku',
 '55\\ TV with premium cable',
 '55\\ TV with standard cable',
 '57\\ HDTV with Roku',
 '60\\ HDTV with Amazon Prime Video',
 '65\\ HDTV with Amazon Prime Video',
 '65\\ HDTV with Chromecast',
 '65\\ HDTV with Netflix',
 '65\\ HDTV with standard cable',
 '70\\ HDTV with Amazon Prime Video',
 '75\\ HDTV with Amazon Prime Video',
 '75\\ HDTV with Roku',
 '75\\ TV',
 ':) body soap',
 ':) conditioner',
 'Aesop body soap',
 'Air conditioning',
 'Alaffia',
 'Amana refrigerator',
 'Amazon  conditioner',
 'Amazon Prime Video',
 'Amazon body soap',
 'Apple TV',
 'Audio-Technica Record Player  sound system with Bluetooth and aux',
 'Avanti stainless steel electric stove',
 'Avanti stainless steel oven',
 'Aveeno body soap',
 'BBQ grill',
 'Baby bath',
 'Baby monitor',
 'Baby safety gates',
 'Babysitter recommendations',
 'Backyard',
 'Baking sheet',
 'Barbecue utensils',
 'Bath & Body Works body soap',
 'Bath & Body Works conditioner',
 'Bath & Body Works shampoo',
 'Bathtub',
 'Beach essentials',
 'Beachfront',
 'Bed linens',
 'Bed sheets and pillows',
 'Beekman Hotel conditioner',
 'Beekman body soap',
 'Beekman conditioner',
 'Beekman shampoo',
 'Bidet',
 'Bikes',
 'Bluetooth sound system',
 'Bluetooth speaker',
 'Bluetooth speaker Bluetooth sound system',
 'Board games',
 'Boat slip',
 'Body soap',
 'Bosch refrigerator',
 'Bosch stainless steel oven',
 'Bose sound system with Bluetooth and aux',
 'Both bar soaps and body wash are available for you. body soap',
 'Bread maker',
 'Breakfast',
 'Breakfast buffet available \\u2014 $19 per person per day',
 'Building staff',
 'Cable TV',
 'Cambridge Soundworks / Doss (two small units) sound system with Bluetooth and aux',
 'Carbon monoxide alarm',
 'Ceiling fan',
 'Central AC  conditioner',
 'Central air conditioning',
 'Central heating',
 'Changing table',
 'Children\\u2019s books and toys',
 'Children\\u2019s books and toys for ages 0-2 years old',
 'Children\\u2019s books and toys for ages 0-2 years old and 2-5 years old',
 'Children\\u2019s books and toys for ages 2-5 years old',
 'Children\\u2019s books and toys for ages 2-5 years old and 5-10 years old',
 'Children\\u2019s books and toys for ages 5-10 years old and 10+ years old',
 'Children\\u2019s dinnerware',
 'Chromecast',
 'Cleaning before checkout',
 'Cleaning products',
 'Clothing storage',
 'Clothing storage: closet',
 'Clothing storage: closet and dresser',
 'Clothing storage: closet and walk-in closet',
 'Clothing storage: closet and wardrobe',
 'Clothing storage: dresser',
 'Clothing storage: dresser and closet',
 'Clothing storage: dresser and walk-in closet',
 'Clothing storage: dresser and wardrobe',
 'Clothing storage: walk-in closet',
 'Clothing storage: walk-in closet and closet',
 'Clothing storage: walk-in closet and dresser',
 'Clothing storage: wardrobe',
 'Clothing storage: wardrobe and closet',
 'Clothing storage: wardrobe and dresser',
 'Coffee maker',
 'Complimentary continental breakfast',
 'Concierge',
 'Conditioner',
 'Connect your device via Bluetooth sound system',
 'Cooking basics',
 'Crib',
 'DOVE Usually body soap',
 'Dedicated workspace',
 'Dedicated workspace: desk',
 'Dedicated workspace: desk and office chair',
 'Dedicated workspace: desk and table',
 'Dedicated workspace: monitor',
 'Dedicated workspace: monitor and desk',
 'Dedicated workspace: office chair',
 'Dedicated workspace: office chair and desk',
 'Dedicated workspace: office chair and table',
 'Dedicated workspace: table',
 'Dedicated workspace: table and desk',
 'Dedicated workspace: table and office chair',
 'Dial Soap Bars body soap',
 'Dining table',
 'Dishes and silverware',
 'Dishwasher',
 'Dove',
 'Dove body soap',
 'Dove conditioner',
 'Dove shampoo',
 'Dryer',
 'Dryer \\u2013 In building',
 'Dryer \\u2013\\u00a0In unit',
 'Drying rack for clothing',
 'EV charger',
 'Ecco sound system',
 'Eco bar soap body soap',
 'Electric stove',
 'Elevator',
 'Essentials',
 'Ethernet connection',
 'Extra pillows and blankets',
 'Fast wifi \\u2013 100 Mbps',
 'Fast wifi \\u2013 102 Mbps',
 'Fast wifi \\u2013 103 Mbps',
 'Fast wifi \\u2013 121 Mbps',
 'Fast wifi \\u2013 157 Mbps',
 'Fast wifi \\u2013 191 Mbps',
 'Fast wifi \\u2013 220 Mbps',
 'Fast wifi \\u2013 276 Mbps',
 'Fast wifi \\u2013 302 Mbps',
 'Fast wifi \\u2013 395 Mbps',
 'Fast wifi \\u2013 409 Mbps',
 'Fast wifi \\u2013 445 Mbps',
 'Fast wifi \\u2013 51 Mbps',
 'Fast wifi \\u2013 54 Mbps',
 'Fast wifi \\u2013 55 Mbps',
 'Fast wifi \\u2013 58 Mbps',
 'Fast wifi \\u2013 59 Mbps',
 'Fast wifi \\u2013 63 Mbps',
 'Fast wifi \\u2013 68 Mbps',
 'Fast wifi \\u2013 78 Mbps',
 'Fast wifi \\u2013 86 Mbps',
 'Fast wifi \\u2013 93 Mbps',
 'Fast wifi \\u2013 95 Mbps',
 'Fast wifi \\u2013 96 Mbps',
 'Fast wifi \\u2013 97 Mbps',
 'Fast wifi \\u2013 98 Mbps',
 'Fenced garden or backyard',
 'Fire extinguisher',
 'Fire pit',
 'Fireplace guards',
 'First aid kit',
 'Free carport on premises \\u2013 1 space',
 'Free driveway parking on premises',
 'Free driveway parking on premises \\u2013 1 space',
 'Free driveway parking on premises \\u2013 2 spaces',
 'Free dryer',
 'Free dryer \\u2013 In building',
 'Free dryer \\u2013 In unit',
 'Free parking on premises',
 'Free parking on premises \\u2013 1 space',
 'Free residential garage on premises',
 'Free residential garage on premises \\u2013 1 space',
 'Free street parking',
 'Free washer',
 'Free washer \\u2013 In building',
 'Free washer \\u2013 In unit',
 'Free wifi',
 'Freezer',
 'Freshscent body soap',
 'Freshscent shampoo',
 'GE 20.2 CU FT refrigerator',
 'GE Profile refrigerator',
 'GE Profile stainless steel gas stove',
 'GE Profile stainless steel oven',
 'GE gas stove',
 'GE oven',
 'GE refrigerator',
 'Gaggenau stainless steel gas stove',
 'Game console',
 'Game console: Nintendo Wii',
 'Game console: PS2',
 'Game console: PS4',
 'Game console: PS5',
 'Game room',
 'Gas oven',
 'Gas stove',
 'Gel body soap',
 'Gel shampoo',
 'Generic conditioner',
 'Gilchrist And Saomes body soap',
 'Gilchrist and Soames body soap',
 'Gilchrist and Somes body soap',
 'Google Home Bluetooth sound system',
 'Gym',
 'HBO Max',
 'HDTV',
 'HDTV with Amazon Prime Video',
 'HDTV with Apple TV',
 'HDTV with Chromecast',
 'HDTV with HBO Max',
 'HDTV with Netflix',
 'HDTV with Roku',
 'HDTV with premium cable',
 'HDTV with standard cable',
 'Hair dryer',
 'Hangers',
 'Heating',
 'High chair',
 'Honest body soap',
 'Host greets you',
 'Hot tub',
 'Hot water',
 'Hot water kettle',
 'House bikes',
 'Housekeeping',
 'Ikea gas stove',
 'Ikea oven',
 'Indoor fireplace',
 'Indoor pool',
 'Induction stove',
 'Infuse - Made from white tea and coconut conditioner',
 'Infuse made from white tea and coconut body soap',
 'Infuse- made from white tea And coconut shampoo',
 'Irish Spring Shower gel/Dial liquid hand soap/Tulip body soap body soap',
 'Iron',
 'Ivory body soap',
 'KEF sound system with Bluetooth and aux',
 'Kayak',
 'Kenmore electric stove',
 'Kenmore oven',
 'Kenmore stainless steel electric stove',
 'Kenmore stainless steel oven',
 'Keurig coffee machine',
 'Keypad',
 "Kiehl's body soap",
 "Kiehl's conditioner",
 'Kirkland body soap',
 'Kirkland conditioner',
 'Kirkland shampoo',
 'Kitchen',
 'KitchenAid refrigerator',
 'KitchenAid stainless steel gas stove',
 'KitchenAid stainless steel oven',
 'Kitchenaid refrigerator',
 'Kitchenaid stainless steel gas stove',
 'Kitchenaid stainless steel oven',
 'Kitchenette',
 'LG New as of 3/21 stainless steel oven',
 'LG refrigerator',
 'Lake access',
 'Laundromat nearby',
 'Laundry services',
 'Limited housekeeping \\u2014',
 'Limited housekeeping \\u2014 on request',
 'Liquid body soap',
 'Liquid conditioner',
 'Liquid shampoo',
 'Lock on bedroom door',
 'Lockbox',
 'Long term stays allowed',
 'Loreal conditioner',
 'Luggage dropoff allowed',
 'MIlk',
 'Maxxam conditioner',
 'Maxxam shampoo',
 'Microwave',
 'Milk conditioner',
 'Mini fridge',
 'Mosquito net',
 'Nespresso machine',
 'Netflix',
 'Neutrogena or Kirkland body soap',
 'Nexus or Kirkland shampoo',
 'No stove but has a hot plate electric stove',
 'Not sure  body soap',
 'Olay Body Wash with Retinol',
 'Olay body soap',
 'Onsite restaurant \\u2014 Cosmica',
 'Onsite restaurant \\u2014 canteenM-open 24/7 Open 24/7',
 'Organic  body soap',
 'Organic 365% and EVERYBODY shampoo',
 'Organic conditioner',
 'Organic shampoo',
 'Oster Toaster Oven oven',
 'Outdoor dining area',
 'Outdoor furniture',
 'Outdoor shower',
 'Outlet covers',
 'Oven',
 'PANTENE Usually conditioner',
 'PANTENE Usually shampoo',
 'Pack \\u2019n play/Travel crib',
 'Paid dryer',
 'Paid dryer \\u2013 In building',
 'Paid dryer \\u2013 In unit',
 'Paid parking garage off premises',
 'Paid parking garage on premises',
 'Paid parking garage on premises \\u2013 1 space',
 'Paid parking lot off premises',
 'Paid parking lot on premises',
 'Paid parking lot on premises \\u2013 1 space',
 'Paid parking lot on premises \\u2013 2 spaces',
 'Paid parking lot on premises \\u2013 3 spaces',
 'Paid parking lot on premises \\u2013 4 spaces',
 'Paid parking lot on premises \\u2013 6 spaces',
 'Paid parking off premises',
 'Paid parking on premises',
 'Paid parking on premises \\u2013 1 space',
 'Paid street parking off premises',
 'Paid valet parking on premises',
 'Paid washer',
 'Paid washer \\u2013 In building',
 'Paid washer \\u2013 In unit',
 'Panasonic sound system',
 'Pantene  conditioner',
 'Pantene conditioner',
 'Pantene shampoo',
 'Patio or balcony',
 'Pets allowed',
 'Piano',
 'Ping pong table',
 'Pocket wifi',
 'Polk Bluetooth sound system',
 'Pool',
 'Pool table',
 'Portable air conditioning',
 'Portable fans',
 'Portable heater',
 'Pour-over coffee',
 'Private entrance',
 'Private fenced garden or backyard',
 'Private garden or backyard',
 'Private gym in building',
 'Private hot tub',
 'Private outdoor lap pool',
 'Private patio or balcony',
 'Private pool',
 'Private sauna',
 'Purchased in 2020. refrigerator',
 'Radiant heating',
 'Radio and CD player sound system',
 'Record player',
 'Refrigerator',
 'Rice maker',
 'Rikoko conditioner',
 'Roku',
 'Room-darkening shades',
 'SONOS Bluetooth sound system',
 'Safe',
 'Samsung  gas stove',
 'Samsung  stainless steel gas stove',
 'Samsung  stainless steel oven',
 'Samsung french door refrigerator',
 'Samsung oven',
 'Samsung refrigerator',
 'Samsung stainless steel gas stove',
 'Samsung stainless steel oven',
 'Security cameras on property',
 'Self-parking \\u2014 $42/stay',
 'Shampoo',
 'Shared fenced garden or backyard',
 'Shared garden or backyard',
 'Shared gym in building',
 'Shared gym nearby',
 'Shared hot tub',
 'Shared outdoor heated saltwater pool',
 'Shared outdoor pool',
 'Shared patio or balcony',
 'Shared pool',
 'Shared refrigerator',
 'Shower gel',
 'Single level home',
 'Ski-in/Ski-out',
 'Slippers',
 'Small fridge in room. Larger fridge in kitchenette  refrigerator',
 'Smart lock',
 'Smoke alarm',
 'SoapBox Sea Minerals & Blue Iris body soap',
 'SoapBox Tea Tree Clean & Purify conditioner',
 'SoapBox Tea Tree Clean & Purify shampoo',
 'Softsoap body soap',
 'Sono sound system',
 'Sonos  sound system',
 'Sonos Bluetooth sound system',
 'Sonos and Amazon Tap Bluetooth sound system',
 'Sonos sound system',
 'Sonos sound system with Bluetooth and aux',
 'Sonos wifi sound system',
 'Sound system',
 'Sound system with Bluetooth and aux',
 'Sound system with aux',
 'Stainless steel electric stove',
 'Stainless steel gas stove',
 'Stainless steel oven',
 'Stainless steel refrigerator',
 'Stainless steel stove',
 'Stove',
 'Suave body soap',
 'TRESemm\\u00e9 conditioner',
 'TRESemm\\u00e9 shampoo',
 'TV',
 'TV with Amazon Prime Video',
 'TV with Apple TV',
 'TV with Chromecast',
 'TV with HBO Max',
 'TV with Netflix',
 'TV with Roku',
 'TV with standard cable',
 'Table corner guards',
 'Teenage Engineering Bluetooth sound system',
 'There are two induction burners with pots and pans. induction stove',
 'There is a 24\\ Blomberg fridge/freezer in the unit. refrigerator',
 'Thermador  stainless steel electric stove',
 'Toaster',
 'Toiletries',
 'Toshiba with Boss Speakers sound system with Bluetooth and aux',
 'Trash compactor',
 'Tresemme  conditioner',
 'Tresemme shampoo',
 'Varies conditioner',
 'Various body soap',
 'Various conditioner',
 'WHIRLPOOL refrigerator',
 'WHIRPOOL refrigerator',
 'WHRILPOOL refrigerator',
 'Washer',
 'Washer \\u2013\\u00a0In building',
 'Washer \\u2013\\u00a0In unit',
 'Waterfront',
 'Whirlpool  refrigerator',
 'Whirlpool refrigerator',
 'Whirlpool stainless steel oven',
 'Whole Foods body soap',
 'Wifi',
 'William Roam toiletries',
 'Window AC unit',
 'Window guards',
 'Wine Mini Fridge (For Beverages Only) refrigerator',
 'Wine glasses',
 'Your choice 100% Organic or Skin Sensitive no scent - MILK Brand body soap',
 'and 10+ years old',
 'and 5-10 years old',
 'and Axe Body Wash body soap',
 'and closet',
 'and desk',
 'and dresser',
 'and goodies for sharing. Our baristas will make you a perfect coffee all day (and night) long',
 'and in the evenings they transform into expert mixologists!',
 'and monitor',
 'and office chair',
 'and table',
 'and walk-in closet',
 'bose sound system with Bluetooth and aux',
 'canteenM is our casual and cosy kitchen',
 'closet',
 'desk',
 'dove  conditioner',
 'dove body soap',
 'dresser',
 'google assistant sound system',
 'healthy snacks',
 'hypoallergenic body soap',
 'linens',
 'mini fridge refrigerator',
 'minifridge refrigerator',
 'monitor',
 'offering a choice of homemade meals',
 'office chair',
 'premium cable',
 'something nice! :) body soap',
 'standard cable',
 'table',
 'toiletries',
 'trader joes conditioner',
 'under counter refrigerator',
 'vegan and cruelty-free body soap',
 'vegan and cruelty-free conditioner',
 'vegan and cruelty-free shampoo',
 'walk-in closet',
 'wardrobe',
 'whirlpool  oven'}

boston3.loc[boston3['amenities'].str.contains('Air conditioning|Central air conditioning|Portable air conditioning'), 'air_conditioning'] = 1
boston3.loc[boston3['amenities'].str.contains('Gym|24-hour fitness center|Private gym in building|Shared gym in building|Shared gym nearby'), 'gym'] = 1
boston3.loc[boston3['amenities'].str.contains('Apple TV|Game console|Game console: Nintendo Wii|Game console: PS2|Game console: PS4|Game console: PS5|Netflix|32\\ HDTV|32\\ HDTV with|32\\ HDTV with Amazon Prime Video|32\\ HDTV with Netflix|32\\ HDTV with standard cable|36\\ HDTV with Apple|37\\ HDTV with Amazon Prime Video|40\\ HDTV with Roku|40\\ HDTV with standard cable|40\\ TV|42\\ HDTV|42\\ HDTV with Amazon Prime Video|42\\ HDTV with Roku|45\\ HDTV with premium cable|48\\ HDTV|50\\ HDTV with Roku|50\\ HDTV with standard cable|52\\ HDTV|54\\ HDTV with Netflix|55\\ HDTV|55\\ HDTV with Apple TV|55\\ HDTV with Netflix|65\\ HDTV with Amazon Prime Video|65\\ HDTV with Netflix|75\\ TV|Amazon Prime Video|Apple TV| Cable TV|HBO Max|HDTV|HDTV with Amazon Prime Video|HDTV with Apple TV|HDTV with HBO Max|HDTV with Netflix|HDTV with Roku|HDTV with premium cable|HDTV with standard cable| Netflix|TV|TV with Amazon Prime Video|TV with Chromecast|TV with HBO Max|TV with Netflix|TV with Roku|TV with standard cable'), 'tv'] = 1
boston3.loc[boston3['amenities'].str.contains('Avanti stainless steel electric stove|Avanti stainless steel oven|Ikea oven|Kenmore oven|Bosch stainless steel oven'), 'oven'] = 1
boston3.loc[boston3['amenities'].str.contains('Electric stove|Gas stove|Ikea gas stove|Kenmore electric stove|Kenmore stainless steel electric stove|Bosch stainless steel gas stove'), 'stove'] = 1
boston3.loc[boston3['amenities'].str.contains('BBQ grill|Fire pit|Barbecue utensils'), 'bbq'] = 1
boston3.loc[boston3['amenities'].str.contains('balcony|Patio or balcony|Patio'),'nature_and_views'] = 1
boston3.loc[boston3['amenities'].str.contains('Beach view|Beachfront|Lake access|Mountain view|Ski-in/Ski-out|Waterfront|Lake access|Beachfront||Ski-in/Ski-out'), 'nature_and_views'] = 1
boston3.loc[boston3['amenities'].str.contains('Bed linens|Bed sheets and pillows|linens'), 'bed_linen'] = 1
boston3.loc[boston3['amenities'].str.contains('Breakfast|Complimentary continental breakfast|Continental breakfast available \\u2014 $14 per person per day'), 'breakfast'] = 1
boston3.loc[boston3['amenities'].str.contains('Coffee maker|Espresso machine|Coffee maker|Keurig coffee machine|Nespresso machine|Pour-over coffee'), 'coffee_machine'] = 1
boston3.loc[boston3['amenities'].str.contains('Cooking basics'), 'cooking_basics'] = 1
boston3.loc[boston3['amenities'].str.contains('Dryer|Dryer \\u2013 In building|Dryer \\u2013\\u00a0In unit|Drying rack for clothing|Dishwasher|Free washer\\u2013In building|Free washer\\u2013In unit|Paid washer\\u2013In building|WasherWasher \\u2013\\u00a0In building|Washer \\u2013\\u00a0In unit'), 'white_goods'] = 1
boston3.loc[boston3['amenities'].str.contains('Elevator'), 'elevator'] = 1
boston3.loc[boston3['amenities'].str.contains('Children\\u2019s books and toys|Children\\u2019s books and toys for ages 0-2 years old|Children\\u2019s books and toys for ages 0-2 years old and 2-5 years old|Children\\u2019s dinnerware|Baby bath|Baby monitor|Baby safety gates|Babysitter recommendations'), 'child_friendly'] = 1
boston3.loc[boston3['amenities'].str.contains('Free driveway parking on premises|Free driveway parking on premises \\u2013 1 space|Free driveway parking on premises \\u2013 2 spaces| Free parking garage on premises|Free parking on premises|Free street parking|Paid parking garage off premises|Paid parking garage off premises \\u2013 $30 per day|Paid parking garage off premises \\u2013 $32 per day|Paid parking garage off premises \\u2013 $35 per day|Paid parking garage off premises \\u2013 $40 per day|Paid parking garage off premises \\u2013 $45 per day|Paid parking garage off premises \\u2013 $48 per day|Paid parking garage off premises \\u2013 $50 per day|Paid parking garage on premises|Paid parking garage on premises \\u2013 1 space|Paid parking lot off premises|Paid parking lot off premises \\u2013 $35 per day|Paid parking lot on premises|Paid parking lot on premises \\u2013 $50 per day|Paid parking lot on premises \\u2013 $65 per day|Paid parking lot on premises \\u2013 1 space|Paid parking lot on premises \\u2013 2 spaces|Paid parking off premises|Paid parking on premises|Paid street parking off premises'), 'parking'] = 1
boston3.loc[boston3['amenities'].str.contains('Private fenced garden or backyard|Private garden or backyard|Garden|Outdoor|Sun loungers|Terrace|Backyard'), 'outdoor_space'] = 1
boston3.loc[boston3['amenities'].str.contains('Host greets you'), 'host_greeting'] = 1
boston3.loc[boston3['amenities'].str.contains('Private hot tub|Shared outdoor heated saltwater pool|Shared outdoor pool|Shared pool|Hot tub|Jetted tub|hot tub|Sauna|Pool|pool'), 'hot_tub_sauna_or_pool'] = 1
boston3.loc[boston3['amenities'].str.contains('Internet|Pocket wifi|Wifi|Pocket wifi|Wifi|Wifi \\u2013 100 Mbps|Wifi \\u2013 1000 Mbps|Wifi \\u2013 15 Mbps|Wifi \\u2013 150 Mbps|Wifi \\u2013 180 Mbps|Wifi \\u2013 200 Mbps|Wifi \\u2013 250 Mbps|Wifi \\u2013 363 Mbps|Wifi \\u2013 60 Mbps|Wifi \\u2013 600 Mbps|Wifi \\u2013 70 Mbps|Wifi \\u2013 90 Mbps|Free wifi'), 'internet'] = 1
boston3.loc[boston3['amenities'].str.contains('Long term stays allowed'), 'long_term_stays'] = 1
boston3.loc[boston3['amenities'].str.contains('Private entrance'), 'private_entrance'] = 1
boston3.loc[boston3['amenities'].str.contains('Pets|pet|Cat(s)|Dog(s)|Pets allowed'), 'pets_allowed'] = 1
boston3.loc[boston3['amenities'].str.contains('Safe|Security system|Security cameras on property'), 'secure'] = 1
boston3.loc[boston3['amenities'].str.contains('Self check_in'), 'self_check_in'] = 1
boston3.loc[boston3['amenities'].str.contains('Smoking allowed'), 'smoking_allowed'] = 1
boston3.loc[boston3['amenities'].str.contains('Step-free access|Wheelchair|Accessible'), 'accessible'] = 1
boston3.loc[boston3['amenities'].str.contains('Suitable for events'), 'event_suitable'] = 1 
boston3.loc[boston3['amenities'].str.contains('Aveeno body soap|Bathtub|Beekman Hotel conditioner|Beekman body soap|Beekman conditioner|Beekman shampoo|Bath & Body Works body soap|Bath & Body Works body soap|Bath & Body Works conditioner|Bath & Body Works shampoo|Bath & Body Works. body soap|Body soap//Both bar soaps and body wash are available for you. body soap|DOVE Usually body soap|Dove body soap|Dove conditioner|Dove shampoo|Eco body soap|Gilchrist And Saomes body soap|Gilchrist and Soames body soap|Gilchrist and Somes body soap|Irish Spring Shower gel|Dial liquid hand soap|Tulip body soap body soap|Maxxam conditioner|Maxxam shampoo|PANTENE Usually conditioner|PANTENE Usually shampoo|TRESemm\\u00e9 conditioner|TRESemm\\u00e9 shampoo|Loreal conditioner'), 'toiletries'] = 1 
boston3.loc[boston3['amenities'].str.contains('Sonos Bluetooth sound system|Sonos and Amazon Tap Bluetooth sound system|bose sound system with Bluetooth and aux|Bluetooth sound system|Bluetooth speaker|Bluetooth speaker Bluetooth sound system|Bose Sound Touch with Bluetooth sound system with Bluetooth and aux'), 'sound_system'] = 1
boston3.loc[boston3['amenities'].str.contains('Dedicated workspace|Dedicated workspace: desk|Dedicated workspace: desk and office chair|Dedicated workspace: desk and table|Dedicated workspace: monitor and desk|Dedicated workspace: office chair and desk|Dedicated workspace: table|Dedicated workspace: table and desk|Dedicated workspace: table and office chair'), 'workspace'] = 1
boston3.loc[boston3['amenities'].str.contains('Freezer|GE refrigerator| LG refrigerator|Purchased in 2020. Refrigerator|Samsung refrigerator|WHIRLPOOL refrigerator|WHRILPOOL refrigerator|Mini fridge|KitchenAid refrigerator|Kitchenaid refrigerator'), 'refrigerator'] = 1
boston3.loc[boston3['amenities'].str.contains('Onsite restaurant \\u2014 Cosmica|Onsite restaurant \\u2014 canteenM-open 24/7 Open 24/7'), 'canteen'] = 1

<ipython-input-21-90ad71994ea5>:23: UserWarning: This pattern is interpreted as a regular expression, and has match groups. To actually get the groups, use str.extract.
  boston3.loc[boston3['amenities'].str.contains('Pets|pet|Cat(s)|Dog(s)|Pets allowed'), 'pets_allowed'] = 1

replacecols = boston3.iloc[:,0:].columns
boston3[replacecols] = boston3[replacecols].fillna(0)
nonessential_amenities = []
for col in boston3.iloc[:,17:].columns:
    if boston3[col].sum() < len(boston3)/10:
        nonessential_amenities.append(col)
boston3.drop(nonessential_amenities, axis=1, inplace=True)
boston3.drop('amenities', axis=1, inplace=True)

boston3.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 107389 entries, 0 to 107388
Data columns (total 34 columns):
 #   Column                          Non-Null Count   Dtype  
---  ------                          --------------   -----  
 0   name                            107389 non-null  object 
 1   host_id                         107389 non-null  int64  
 2   neighbourhood                   107389 non-null  object 
 3   latitude                        107389 non-null  float64
 4   longitude                       107389 non-null  float64
 5   room_type                       107389 non-null  object 
 6   price                           107389 non-null  int64  
 7   minimum_nights                  107389 non-null  int64  
 8   number_of_reviews               107389 non-null  int64  
 9   reviews_per_month               107389 non-null  float64
 10  calculated_host_listings_count  107389 non-null  int64  
 11  availability_365                107389 non-null  int64  
 12  number_of_reviews_ltm           107389 non-null  int64  
 13  accommodates                    107389 non-null  int64  
 14  bedrooms                        107389 non-null  float64
 15  beds                            107389 non-null  float64
 16  air_conditioning                107389 non-null  float64
 17  gym                             107389 non-null  float64
 18  tv                              107389 non-null  float64
 19  bbq                             107389 non-null  float64
 20  nature_and_views                107389 non-null  float64
 21  bed_linen                       107389 non-null  float64
 22  coffee_machine                  107389 non-null  float64
 23  cooking_basics                  107389 non-null  float64
 24  white_goods                     107389 non-null  float64
 25  elevator                        107389 non-null  float64
 26  parking                         107389 non-null  float64
 27  hot_tub_sauna_or_pool           107389 non-null  float64
 28  internet                        107389 non-null  float64
 29  long_term_stays                 107389 non-null  float64
 30  private_entrance                107389 non-null  float64
 31  toiletries                      107389 non-null  float64
 32  workspace                       107389 non-null  float64
 33  refrigerator                    107389 non-null  float64
dtypes: float64(23), int64(8), object(3)
memory usage: 28.7+ MB

Data Visualization

boston3=boston3.drop(['nature_and_views'],axis=1)

price_corr = boston3.corr()['price'].sort_values(ascending = False).head(6)[1:]
price_corr

private_entrance    0.457659
coffee_machine      0.415872
accommodates        0.413670
bed_linen           0.406062
gym                 0.403006
Name: price, dtype: float64

### Features that influence Price
price_corr.plot(kind = 'bar', color = 'g');
plt.xticks(rotation=45);

boston3.head()

	name	host_id	neighbourhood	latitude	longitude	room_type	price	minimum_nights	number_of_reviews	reviews_per_month	...	white_goods	elevator	parking	hot_tub_sauna_or_pool	internet	long_term_stays	private_entrance	toiletries	workspace	refrigerator
0	HARBORSIDE-Walk to subway	4804	East Boston	42.36413	-71.02991	Entire home/apt	125	32	21	0.27	...	1.0	0.0	1.0	0.0	1.0	1.0	0.0	0.0	1.0	0.0
1	** Private! Minutes to center!**	8229	Roxbury	42.32844	-71.09581	Entire home/apt	99	3	110	0.71	...	1.0	0.0	1.0	0.0	1.0	1.0	1.0	1.0	1.0	0.0
2	** Private! Minutes to center!**	8229	Roxbury	42.32844	-71.09581	Entire home/apt	99	3	110	0.71	...	1.0	0.0	1.0	0.0	1.0	1.0	0.0	0.0	1.0	0.0
3	** Private! Minutes to center!**	8229	Roxbury	42.32844	-71.09581	Entire home/apt	99	3	110	0.71	...	1.0	0.0	1.0	0.0	1.0	1.0	1.0	0.0	1.0	0.0
4	** Private! Minutes to center!**	8229	Roxbury	42.32844	-71.09581	Entire home/apt	99	3	110	0.71	...	1.0	0.0	1.0	0.0	1.0	1.0	0.0	1.0	1.0	0.0

5 rows × 33 columns

### Barplot to find the price of different neighbourhoods in Boston

plt.figure(figsize=(15,10))
plt.xticks(rotation=50)
listings = boston3.sort_values(by = 'price')
sb.barplot(x='neighbourhood', y= 'price',palette= "bright", data = boston3)
plt.xlabel(xlabel='Neighbourhood', fontsize=15)
plt.ylabel(ylabel='price', fontsize=15)

Text(0, 0.5, 'price')

maps = folium.Map(location=[42.361145, -71.057083], zoom_start = 12)

heatmap_data = [[row['latitude'],row['longitude']] for index, row in
             boston3[['latitude', 'longitude']].iterrows()]

htm =  HeatMap(heatmap_data).add_to(maps)

maps

Make this Notebook Trusted to load map: File -> Trust Notebook

from PIL import Image                                                                                

plt.figure(figsize=(13,8))
gx = plt.gca()
boston3.plot(kind='scatter',x='longitude',y='latitude',label='Listings', c='price', ax=gx, cmap=plt.get_cmap('jet'), colorbar=True, alpha=0.3, zorder=4)
plt.legend()
plt.show()

### Plot Map for Distribution of location of Airbnb properties in Boston

import plotly.express as px

fig = px.scatter_mapbox(boston3, lat="latitude", lon="longitude", hover_name="name", hover_data=["neighbourhood"],
                        color_discrete_sequence=["orange"], zoom=10, height=300)
fig.update_layout(mapbox_style="open-street-map")
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()

boston3.describe()

	host_id	latitude	longitude	price	minimum_nights	number_of_reviews	reviews_per_month	calculated_host_listings_count	availability_365	number_of_reviews_ltm	...	white_goods	elevator	parking	hot_tub_sauna_or_pool	internet	long_term_stays	private_entrance	toiletries	workspace	refrigerator
count	1.073890e+05	107389.000000	107389.000000	107389.00000	107389.000000	107389.000000	107389.000000	107389.000000	107389.000000	107389.000000	...	107389.000000	107389.000000	107389.000000	107389.000000	107389.000000	107389.000000	107389.000000	107389.000000	107389.000000	107389.000000
mean	1.758024e+08	42.342998	-71.084340	147.84707	51.843345	11.181425	1.655152	127.744713	275.678971	3.312136	...	0.893630	0.348471	0.345203	0.138971	0.996927	0.993677	0.483709	0.200738	0.330779	0.320973
std	1.221205e+08	0.019351	0.033584	116.54091	36.352791	38.862501	0.960907	71.816091	105.919076	11.022718	...	0.308312	0.476488	0.475437	0.345918	0.055349	0.079265	0.499737	0.400554	0.470496	0.466853
min	4.804000e+03	42.235330	-71.172520	0.00000	1.000000	0.000000	0.010000	1.000000	0.000000	0.000000	...	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000
25%	1.074344e+08	42.335150	-71.101850	50.00000	30.000000	0.000000	1.655152	46.000000	226.000000	0.000000	...	1.000000	0.000000	0.000000	0.000000	1.000000	1.000000	0.000000	0.000000	0.000000	0.000000
50%	1.074344e+08	42.347970	-71.070840	149.00000	32.000000	0.000000	1.655152	168.000000	318.000000	0.000000	...	1.000000	0.000000	0.000000	0.000000	1.000000	1.000000	0.000000	0.000000	0.000000	0.000000
75%	2.978601e+08	42.355180	-71.061640	212.00000	91.000000	1.000000	1.655152	194.000000	364.000000	0.000000	...	1.000000	1.000000	1.000000	0.000000	1.000000	1.000000	1.000000	0.000000	1.000000	1.000000
max	4.349904e+08	42.392790	-70.997810	10000.00000	1000.000000	678.000000	27.170000	194.000000	365.000000	298.000000	...	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000

8 rows × 30 columns

boston3.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 107389 entries, 0 to 107388
Data columns (total 33 columns):
 #   Column                          Non-Null Count   Dtype  
---  ------                          --------------   -----  
 0   name                            107389 non-null  object 
 1   host_id                         107389 non-null  int64  
 2   neighbourhood                   107389 non-null  object 
 3   latitude                        107389 non-null  float64
 4   longitude                       107389 non-null  float64
 5   room_type                       107389 non-null  object 
 6   price                           107389 non-null  int64  
 7   minimum_nights                  107389 non-null  int64  
 8   number_of_reviews               107389 non-null  int64  
 9   reviews_per_month               107389 non-null  float64
 10  calculated_host_listings_count  107389 non-null  int64  
 11  availability_365                107389 non-null  int64  
 12  number_of_reviews_ltm           107389 non-null  int64  
 13  accommodates                    107389 non-null  int64  
 14  bedrooms                        107389 non-null  float64
 15  beds                            107389 non-null  float64
 16  air_conditioning                107389 non-null  float64
 17  gym                             107389 non-null  float64
 18  tv                              107389 non-null  float64
 19  bbq                             107389 non-null  float64
 20  bed_linen                       107389 non-null  float64
 21  coffee_machine                  107389 non-null  float64
 22  cooking_basics                  107389 non-null  float64
 23  white_goods                     107389 non-null  float64
 24  elevator                        107389 non-null  float64
 25  parking                         107389 non-null  float64
 26  hot_tub_sauna_or_pool           107389 non-null  float64
 27  internet                        107389 non-null  float64
 28  long_term_stays                 107389 non-null  float64
 29  private_entrance                107389 non-null  float64
 30  toiletries                      107389 non-null  float64
 31  workspace                       107389 non-null  float64
 32  refrigerator                    107389 non-null  float64
dtypes: float64(22), int64(8), object(3)
memory usage: 27.9+ MB

### correlation verify each feature against the target feature

plt.figure(figsize=(12,8))
title = 'Correlation matrix of variables'
sb.heatmap(boston3.corr(), square=True, cmap='YlGnBu')
plt.title(title)
plt.ioff()

<matplotlib.pyplot._IoffContext at 0x7f893dd399a0>

boston3=boston3._get_numeric_data()
boston3=boston3.dropna(axis=0)
boston3=boston3.drop(['latitude','longitude'],axis=1)
boston3=boston3.drop(['number_of_reviews_ltm'],axis=1)
boston3.head()

	host_id	price	minimum_nights	number_of_reviews	reviews_per_month	calculated_host_listings_count	availability_365	accommodates	bedrooms	beds	...	white_goods	elevator	parking	hot_tub_sauna_or_pool	internet	long_term_stays	private_entrance	toiletries	workspace	refrigerator
0	4804	125	32	21	0.27	1	17	2	1.0	1.0	...	1.0	0.0	1.0	0.0	1.0	1.0	0.0	0.0	1.0	0.0
1	8229	99	3	110	0.71	10	75	2	1.0	1.0	...	1.0	0.0	1.0	0.0	1.0	1.0	1.0	1.0	1.0	0.0
2	8229	99	3	110	0.71	10	75	4	1.0	2.0	...	1.0	0.0	1.0	0.0	1.0	1.0	0.0	0.0	1.0	0.0
3	8229	99	3	110	0.71	10	75	6	2.0	3.0	...	1.0	0.0	1.0	0.0	1.0	1.0	1.0	0.0	1.0	0.0
4	8229	99	3	110	0.71	10	75	4	1.0	1.0	...	1.0	0.0	1.0	0.0	1.0	1.0	0.0	1.0	1.0	0.0

5 rows × 27 columns

Multi-Collinearity

def calc_vif(X):

    # Calculating VIF
    vif = pd.DataFrame()
    vif["variables"] = X.columns
    vif["VIF"] = [variance_inflation_factor(X.values, i) for i in range(X.shape[1])]

    return(vif)

x = boston3.iloc[:,:-1]
calc_vif(x)

	variables	VIF
0	host_id	7.131918
1	price	4.364104
2	minimum_nights	7.098761
3	number_of_reviews	1.909629
4	reviews_per_month	6.208965
5	calculated_host_listings_count	17.979208
6	availability_365	11.736312
7	accommodates	17.591002
8	bedrooms	30.443464
9	beds	17.887075
10	air_conditioning	6.963641
11	gym	3.731983
12	tv	11.819420
13	bbq	1.743540
14	bed_linen	9.566801
15	coffee_machine	9.450196
16	cooking_basics	4.028319
17	white_goods	10.950189
18	elevator	3.603388
19	parking	2.176904
20	hot_tub_sauna_or_pool	2.105857
21	internet	135.087303
22	long_term_stays	119.966757
23	private_entrance	6.258084
24	toiletries	1.913979
25	workspace	4.101958

boston3=boston3.drop(['internet'],axis=1)
boston3=boston3.drop(['long_term_stays'],axis=1)
boston3=boston3.drop(['host_id'],axis=1)

boston3.iloc[:,6:20].corr()

	accommodates	bedrooms	beds	air_conditioning	gym	tv	bbq	bed_linen	coffee_machine	cooking_basics	white_goods	elevator	parking	hot_tub_sauna_or_pool
accommodates	1.000000	0.811684	0.791003	0.124014	0.109612	0.253955	0.062040	0.274648	0.323299	0.247697	0.044392	0.132249	-0.086699	0.113075
bedrooms	0.811684	1.000000	0.852407	0.084183	-0.034519	0.109999	-0.041420	0.065092	0.093667	0.158002	0.050702	0.005295	-0.045489	0.019042
beds	0.791003	0.852407	1.000000	0.105523	-0.030464	0.086498	-0.014476	0.055282	0.080691	0.203304	0.030545	-0.031894	0.051400	0.027351
air_conditioning	0.124014	0.084183	0.105523	1.000000	0.006799	0.136304	0.057425	0.228350	0.209567	-0.103488	0.092243	0.223253	-0.011805	-0.209335
gym	0.109612	-0.034519	-0.030464	0.006799	1.000000	0.247503	0.472893	0.365264	0.418689	-0.031880	0.188861	0.600709	-0.299307	0.493113
tv	0.253955	0.109999	0.086498	0.136304	0.247503	1.000000	0.119941	0.439707	0.394103	0.117889	0.156572	0.257625	-0.252522	0.127826
bbq	0.062040	-0.041420	-0.014476	0.057425	0.472893	0.119941	1.000000	0.238824	0.266668	-0.065221	0.123966	0.359661	-0.161452	0.384242
bed_linen	0.274648	0.065092	0.055282	0.228350	0.365264	0.439707	0.238824	1.000000	0.746457	0.069896	0.102198	0.396446	-0.263837	0.198010
coffee_machine	0.323299	0.093667	0.080691	0.209567	0.418689	0.394103	0.266668	0.746457	1.000000	0.177661	0.122758	0.419305	-0.244053	0.210952
cooking_basics	0.247697	0.158002	0.203304	-0.103488	-0.031880	0.117889	-0.065221	0.069896	0.177661	1.000000	-0.016439	-0.245798	0.040522	0.165806
white_goods	0.044392	0.050702	0.030545	0.092243	0.188861	0.156572	0.123966	0.102198	0.122758	-0.016439	1.000000	0.201418	0.026253	0.072773
elevator	0.132249	0.005295	-0.031894	0.223253	0.600709	0.257625	0.359661	0.396446	0.419305	-0.245798	0.201418	1.000000	-0.313354	0.249679
parking	-0.086699	-0.045489	0.051400	-0.011805	-0.299307	-0.252522	-0.161452	-0.263837	-0.244053	0.040522	0.026253	-0.313354	1.000000	-0.200088
hot_tub_sauna_or_pool	0.113075	0.019042	0.027351	-0.209335	0.493113	0.127826	0.384242	0.198010	0.210952	0.165806	0.072773	0.249679	-0.200088	1.000000

boston3 = boston3.loc[boston3['price'] < 5000]

boston3['price'].describe()

count    107382.000000
mean        147.484206
std         106.597849
min           0.000000
25%          50.000000
50%         149.000000
75%         212.000000
max        3999.000000
Name: price, dtype: float64

boston3.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 107382 entries, 0 to 107388
Data columns (total 24 columns):
 #   Column                          Non-Null Count   Dtype  
---  ------                          --------------   -----  
 0   price                           107382 non-null  int64  
 1   minimum_nights                  107382 non-null  int64  
 2   number_of_reviews               107382 non-null  int64  
 3   reviews_per_month               107382 non-null  float64
 4   calculated_host_listings_count  107382 non-null  int64  
 5   availability_365                107382 non-null  int64  
 6   accommodates                    107382 non-null  int64  
 7   bedrooms                        107382 non-null  float64
 8   beds                            107382 non-null  float64
 9   air_conditioning                107382 non-null  float64
 10  gym                             107382 non-null  float64
 11  tv                              107382 non-null  float64
 12  bbq                             107382 non-null  float64
 13  bed_linen                       107382 non-null  float64
 14  coffee_machine                  107382 non-null  float64
 15  cooking_basics                  107382 non-null  float64
 16  white_goods                     107382 non-null  float64
 17  elevator                        107382 non-null  float64
 18  parking                         107382 non-null  float64
 19  hot_tub_sauna_or_pool           107382 non-null  float64
 20  private_entrance                107382 non-null  float64
 21  toiletries                      107382 non-null  float64
 22  workspace                       107382 non-null  float64
 23  refrigerator                    107382 non-null  float64
dtypes: float64(18), int64(6)
memory usage: 20.5 MB

from sklearn.model_selection import train_test_split
x_train, x_test, y_train, y_test = train_test_split(boston3.drop(["price"],axis=1),boston3.price,test_size=0.3,random_state=42)

from ALY6040_Group3 import normalize_data
xscaler,x_processed = normalize_data(x_train)

Decision Tree

from sklearn.tree import DecisionTreeRegressor
from ALY6040_Group3 import train_model,score,predict

model1 = decisionTree = train_model(x_processed,y_train,DecisionTreeRegressor(random_state=42))
start_time = time.time()
e = predict(model1,xscaler,x_test)
o = score(y_test,e)
print(type(model1).__name__,o)
print("Execution time: " + str((time.time() - start_time)) + ' ms')
a,b,c = o
xc=str((time.time() - start_time))

tab = pd.DataFrame({'Actual Values': np.array(y_test).flatten(), 'Decision Tree': e.flatten()}).head(20)
tab.set_index('Actual Values', inplace=True)
tab

DecisionTreeRegressor {'MAE': 11.325720752933316, 'R2': 0.8473190738130825, 'RMSE': 41.55910163033422}
Execution time: 0.020350217819213867 ms

	Decision Tree
Actual Values
34	33.969697
203	201.531915
240	240.000000
128	128.000000
131	131.000000
341	341.000000
40	42.285714
247	247.000000
52	52.000000
34	35.338462
299	299.000000
196	196.000000
177	214.333333
120	120.000000
181	181.000000
205	179.800000
194	187.666667
132	132.000000
277	277.000000
60	60.000000

### Predicted Values vs Actual Values in Decision Tree

plt.figure(figsize=(15,8))
ax1=sb.distplot(y_test,hist=False,color='blue',label='real value')
ax2=sb.distplot(e,hist=False,color='red',label='predicted value')
plt.legend()

/Users/miravparekh/opt/anaconda3/lib/python3.8/site-packages/seaborn/distributions.py:2557: FutureWarning:

`distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `kdeplot` (an axes-level function for kernel density plots).

/Users/miravparekh/opt/anaconda3/lib/python3.8/site-packages/seaborn/distributions.py:2557: FutureWarning:

`distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `kdeplot` (an axes-level function for kernel density plots).

<matplotlib.legend.Legend at 0x7f8889bc4c70>

Feature Importance

### Feature importance plot of Decision Tree

def plot_feature_importance(importance,names,model_type):
    
    feature_importance = np.array(importance)
    feature_names = np.array(names)
    data={'feature_names':feature_names,'feature_importance':feature_importance}
    fi_df = pd.DataFrame(data)
    fi_df.sort_values(by=['feature_importance'], ascending=False,inplace=True)
    plt.figure(figsize=(10,8))
    sb.barplot(x=fi_df['feature_importance'], y=fi_df['feature_names'])
    plt.title(model_type)
    plt.xlabel('FEATURE IMPORTANCE')
    plt.ylabel('FEATURE NAMES')

plot_feature_importance(model1.feature_importances_,x_train.columns,'Decision Tree')

Cross Validation

from sklearn.model_selection import cross_val_score

lm3 = DecisionTreeRegressor()
scores = cross_val_score(lm3, x_train, y_train, scoring='r2', cv=7)
print('scores:',scores)
print('Mean Score',np.mean(scores)) 

scores: [0.87907627 0.83702184 0.82045976 0.86588401 0.89675677 0.77023254
 0.88322284]
Mean Score 0.8503791470722932

Random Forest Regressor

from sklearn.ensemble import RandomForestRegressor

model2 = RandomForestRegressor = train_model(x_processed,y_train,RandomForestRegressor())
start_time = time.time()
u = predict(model2,xscaler,x_test)
r = score(y_test,u)
print(type(model2).__name__,r)
print("Execution time: " + str((time.time() - start_time)) + ' ms')

xc1=str((time.time() - start_time))

tab['Random Forest Regressor'] = np.array(u[:20])
tab

RandomForestRegressor {'MAE': 11.190295587176706, 'R2': 0.9093543087804413, 'RMSE': 32.021917965836806}
Execution time: 0.6434817314147949 ms

	Decision Tree	Random Forest Regressor
Actual Values
34	33.969697	33.925001
203	201.531915	201.523025
240	240.000000	240.000000
128	128.000000	128.000000
131	131.000000	131.180000
341	341.000000	341.000000
40	42.285714	42.209118
247	247.000000	247.000000
52	52.000000	52.000000
34	35.338462	35.395881
299	299.000000	299.000000
196	196.000000	196.000000
177	214.333333	214.460994
120	120.000000	120.000000
181	181.000000	181.050000
205	179.800000	176.761500
194	187.666667	186.478005
132	132.000000	132.000000
277	277.000000	277.000000
60	60.000000	59.980000

### Predicted Values vs Actual Values in Random Forest Regressor

plt.figure(figsize=(15,8))
ax1=sb.distplot(y_test,hist=False,color='blue',label='real value')
ax2=sb.distplot(u,hist=False,color='red',label='predicted value')
plt.legend()

/Users/miravparekh/opt/anaconda3/lib/python3.8/site-packages/seaborn/distributions.py:2557: FutureWarning:

`distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `kdeplot` (an axes-level function for kernel density plots).

/Users/miravparekh/opt/anaconda3/lib/python3.8/site-packages/seaborn/distributions.py:2557: FutureWarning:

`distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `kdeplot` (an axes-level function for kernel density plots).

<matplotlib.legend.Legend at 0x7f893d41f2b0>

Feature Importance

def plot_feature_importance(importance,names,model_type):

    feature_importance = np.array(importance)
    feature_names = np.array(names)
    data={'feature_names':feature_names,'feature_importance':feature_importance}
    fi_df = pd.DataFrame(data)
    fi_df.sort_values(by=['feature_importance'], ascending=False,inplace=True)
    plt.figure(figsize=(10,8))
    sb.barplot(x=fi_df['feature_importance'], y=fi_df['feature_names'])
    plt.title(model_type)
    plt.xlabel('FEATURE IMPORTANCE')
    plt.ylabel('FEATURE NAMES')

plot_feature_importance(model2.feature_importances_,x_train.columns,'RANDOM FOREST')

Hyperparameter Tuning

from sklearn.ensemble import RandomForestRegressor
Rf = RandomForestRegressor(random_state=42)
Rf.fit(x_train, y_train)

RandomForestRegressor(random_state=42)

from sklearn.model_selection import GridSearchCV
param_grid = {
    'bootstrap': [True],
    'max_depth': [80, 85, 90], 
    'max_features': ['auto'],
    'min_samples_leaf': [1],
    'min_samples_split': [2, 4],
    'n_estimators': [780, 800, 820] 
}

grid_search = GridSearchCV(estimator = Rf, 
                           param_grid = param_grid, 
                           cv = 2, n_jobs = -1, verbose = 2, 
                           scoring = 'accuracy')

grid_search.fit(x_train, y_train)
grid_search.best_params_

Fitting 2 folds for each of 18 candidates, totalling 36 fits

---------------------------------------------------------------------------
KeyboardInterrupt                         Traceback (most recent call last)
<ipython-input-55-a52e1defc50d> in <module>
----> 1 grid_search.fit(x_train, y_train)
      2 grid_search.best_params_

~/opt/anaconda3/lib/python3.8/site-packages/sklearn/utils/validation.py in inner_f(*args, **kwargs)
     61             extra_args = len(args) - len(all_args)
     62             if extra_args <= 0:
---> 63                 return f(*args, **kwargs)
     64 
     65             # extra_args > 0

~/opt/anaconda3/lib/python3.8/site-packages/sklearn/model_selection/_search.py in fit(self, X, y, groups, **fit_params)
    839                 return results
    840 
--> 841             self._run_search(evaluate_candidates)
    842 
    843             # multimetric is determined here because in the case of a callable

~/opt/anaconda3/lib/python3.8/site-packages/sklearn/model_selection/_search.py in _run_search(self, evaluate_candidates)
   1286     def _run_search(self, evaluate_candidates):
   1287         """Search all candidates in param_grid"""
-> 1288         evaluate_candidates(ParameterGrid(self.param_grid))
   1289 
   1290 

~/opt/anaconda3/lib/python3.8/site-packages/sklearn/model_selection/_search.py in evaluate_candidates(candidate_params, cv, more_results)
    793                               n_splits, n_candidates, n_candidates * n_splits))
    794 
--> 795                 out = parallel(delayed(_fit_and_score)(clone(base_estimator),
    796                                                        X, y,
    797                                                        train=train, test=test,

~/opt/anaconda3/lib/python3.8/site-packages/joblib/parallel.py in __call__(self, iterable)
   1052 
   1053             with self._backend.retrieval_context():
-> 1054                 self.retrieve()
   1055             # Make sure that we get a last message telling us we are done
   1056             elapsed_time = time.time() - self._start_time

~/opt/anaconda3/lib/python3.8/site-packages/joblib/parallel.py in retrieve(self)
    931             try:
    932                 if getattr(self._backend, 'supports_timeout', False):
--> 933                     self._output.extend(job.get(timeout=self.timeout))
    934                 else:
    935                     self._output.extend(job.get())

~/opt/anaconda3/lib/python3.8/site-packages/joblib/_parallel_backends.py in wrap_future_result(future, timeout)
    540         AsyncResults.get from multiprocessing."""
    541         try:
--> 542             return future.result(timeout=timeout)
    543         except CfTimeoutError as e:
    544             raise TimeoutError from e

~/opt/anaconda3/lib/python3.8/concurrent/futures/_base.py in result(self, timeout)
    432                 return self.__get_result()
    433 
--> 434             self._condition.wait(timeout)
    435 
    436             if self._state in [CANCELLED, CANCELLED_AND_NOTIFIED]:

~/opt/anaconda3/lib/python3.8/threading.py in wait(self, timeout)
    300         try:    # restore state no matter what (e.g., KeyboardInterrupt)
    301             if timeout is None:
--> 302                 waiter.acquire()
    303                 gotit = True
    304             else:

KeyboardInterrupt: 

from sklearn.ensemble import RandomForestRegressor

Rf = RandomForestRegressor(random_state=42, bootstrap= True,
 max_depth= 80,
 max_features= 'auto',
 min_samples_leaf= 1,
 min_samples_split= 2,
 n_estimators= 780)

Rf.fit(x_train, y_train)

ss= Rf.predict(x_test)
ssg = score(y_test,ss)
print(type(Rf).__name__,ssg)

Cross Validation

from sklearn.model_selection import cross_val_score
from sklearn.ensemble import RandomForestRegressor

lm2 = RandomForestRegressor()
scores = cross_val_score(lm2, x_train, y_train, scoring='r2', cv=7)
print('scores:',scores)
print('Mean Score',np.mean(scores))   

XGBoost

import xgboost as xgb

model3 = xgb = train_model(x_processed,y_train,xgb.XGBRegressor())
start_time = time.time()
h = predict(model3,xscaler,x_test)
j = score(y_test,h)
print(type(model3).__name__,j)
print("Execution time: " + str((time.time() - start_time)) + ' ms')

xc2=str((time.time() - start_time))

tab['XGBoost'] = np.array(h[:20])
tab

/Users/miravparekh/opt/anaconda3/lib/python3.8/site-packages/xgboost/compat.py:31: FutureWarning:

pandas.Int64Index is deprecated and will be removed from pandas in a future version. Use pandas.Index with the appropriate dtype instead.

XGBRegressor {'MAE': 15.525294087041559, 'R2': 0.9120283150987354, 'RMSE': 31.546066398182603}
Execution time: 0.055891990661621094 ms

	Decision Tree	Random Forest Regressor	XGBoost
Actual Values
34	33.969697	33.925001	35.097530
203	201.531915	201.523025	196.505692
240	240.000000	240.000000	230.773285
128	128.000000	128.000000	151.425720
131	131.000000	131.180000	131.134521
341	341.000000	341.000000	328.379211
40	42.285714	42.209118	45.962440
247	247.000000	247.000000	233.773773
52	52.000000	52.000000	48.915230
34	35.338462	35.395881	38.652424
299	299.000000	299.000000	310.659882
196	196.000000	196.000000	190.672577
177	214.333333	214.460994	214.501129
120	120.000000	120.000000	130.386276
181	181.000000	181.050000	185.921860
205	179.800000	176.761500	178.372894
194	187.666667	186.478005	193.043060
132	132.000000	132.000000	156.580856
277	277.000000	277.000000	252.791519
60	60.000000	59.980000	67.213402

plt.figure(figsize=(15,8))
ax1=sb.distplot(y_test,hist=False,color='blue',label='real value')
ax2=sb.distplot(u,hist=False,color='red',label='predicted value')
plt.legend()

Feature Importance

def plot_feature_importance(importance,names,model_type):

    feature_importance = np.array(importance)
    feature_names = np.array(names)
    data={'feature_names':feature_names,'feature_importance':feature_importance}
    fi_df = pd.DataFrame(data)
    fi_df.sort_values(by=['feature_importance'], ascending=False,inplace=True)
    plt.figure(figsize=(10,8))
    sb.barplot(x=fi_df['feature_importance'], y=fi_df['feature_names'])
    plt.title(model_type)
    plt.xlabel('FEATURE IMPORTANCE')
    plt.ylabel('FEATURE NAMES')

plot_feature_importance(model3.feature_importances_,x_train.columns,'XGBoost')

Hyperparameter Tuning

import xgboost as xgb
xg = xgb.XGBRegressor(random_state=42)
xg.fit(x_train, y_train)

from sklearn.model_selection import GridSearchCV 
param_grid = {'learning_rate': [0.1, 0.05], 
              'max_depth': [5, 7, 9],
              'n_estimators': [100, 500, 900]}

gs = GridSearchCV(estimator = xg,
                           param_grid = param_grid, 
                           cv = 2, n_jobs = -1, verbose = 2, 
                           scoring = 'accuracy')

gs.fit(x_train, y_train)
gs.best_params_

from xgboost import XGBRegressor
xg = XGBRegressor(random_state = 42, learning_rate = 0.1, max_depth = 5, n_estimators = 100)

xg.fit(x_train,y_train)

xj= xg.predict(x_train)
dsx = score(y_train,xj)
print(type(xg).__name__,dsx)

Cross Validation

from sklearn.model_selection import cross_val_score
import xgboost as xgb
lm1 = xgb.XGBRegressor()
scores = cross_val_score(lm1, x_train, y_train, scoring='r2', cv=7)
print('scores:',scores)
print('Mean Score',np.mean(scores))     

Linear Regression

from sklearn.linear_model import LinearRegression

model4 = linearRegression = train_model(x_processed,y_train,LinearRegression())
start_time = time.time()
w = predict(model4,xscaler,x_test)
p = score(y_test,w)
print(type(model4).__name__,p)
print("Execution time: " + str((time.time() - start_time)) + ' ms')

xc3=str((time.time() - start_time))

tab['Linear Regression'] = np.array(w[:20])
tab

LinearRegression {'MAE': 41.337713933599446, 'R2': 0.4810204575904966, 'RMSE': 76.6211873143184}
Execution time: 0.013142108917236328 ms

	Decision Tree	Random Forest Regressor	XGBoost	Linear Regression
Actual Values
34	33.969697	33.925001	35.097530	59.999143
203	201.531915	201.523025	196.505692	207.851565
240	240.000000	240.000000	230.773285	203.099159
128	128.000000	128.000000	151.425720	193.862797
131	131.000000	131.180000	131.134521	154.622943
341	341.000000	341.000000	328.379211	206.177947
40	42.285714	42.209118	45.962440	47.359935
247	247.000000	247.000000	233.773773	210.418266
52	52.000000	52.000000	48.915230	70.707581
34	35.338462	35.395881	38.652424	103.220043
299	299.000000	299.000000	310.659882	255.795042
196	196.000000	196.000000	190.672577	206.727563
177	214.333333	214.460994	214.501129	210.148438
120	120.000000	120.000000	130.386276	187.213011
181	181.000000	181.050000	185.921860	199.358890
205	179.800000	176.761500	178.372894	205.837079
194	187.666667	186.478005	193.043060	203.354411
132	132.000000	132.000000	156.580856	198.270749
277	277.000000	277.000000	252.791519	172.610189
60	60.000000	59.980000	67.213402	143.253494

Cross Validation

from sklearn.model_selection import cross_val_score

lm = LinearRegression()
scores = cross_val_score(lm, x_train, y_train, scoring='r2', cv=7)
print('scores:',scores)
print('Mean Score',np.mean(scores))   

plt.figure(figsize=(15,8))
ax1=sb.distplot(y_test,hist=False,color='blue',label='real value')
ax2=sb.distplot(u,hist=False,color='red',label='predicted value')
plt.legend()

Feature Importance

importance = model4.coef_
for i,v in enumerate(importance):
    print('Feature: %0d, Score: %.5f' % (i,v))
    plt.xlabel('FEATURE IMPORTANCE')
    plt.ylabel('FEATURE NAMES')
plot_feature_importance([x for x in range(len(importance))],importance,'Linear Regression')

Regularization

# Creating a linear regression model with Elastic net
re = ElasticNet(alpha=1.0,l1_ratio=0.5)
re.fit(boston3.drop(["price"],axis=1),boston3.price)
y_pred=re.predict(x_test)

print("RMSE: %.3f" % mean_squared_error(y_test, y_pred))
print("R2: %.3f" % r2_score(y_test, y_pred))

import statsmodels.formula.api as smf

#fit regression model
fit = smf.ols('price ~ host_id+minimum_nights+number_of_reviews+reviews_per_month+calculated_host_listings_count+availability_365+accommodates+bedrooms+beds+air_conditioning+gym+tv+bbq+bed_linen+coffee_machine+cooking_basics+white_goods+elevator+parking+hot_tub_sauna_or_pool+private_entrance+toiletries+workspace+refrigerator', data=boston3).fit()

#view model summary
print(fit.summary())

from statsmodels.compat import lzip
import statsmodels.stats.api as sms

#perform Bresuch-Pagan test
names = ['Lagrange multiplier statistic', 'p-value',
        'f-value', 'f p-value']
test = sms.het_breuschpagan(fit.resid, fit.model.exog)

lzip(names, test)

from tabulate import tabulate
 
tab1 = [
    ["Decision Tree", xc],
    ["Random Forest Regressor", xc1],
    ["XGBoost", xc2],
    ["Linear Regression", xc3]
]
 
head = ["Model", "Execution Time"]
print(tabulate(tab1, headers=head, tablefmt="grid"))

from tabulate import tabulate
 
xz=o['RMSE']
xz1=r['RMSE']
xz2=dsx['RMSE']
xz3=p['RMSE']

xxz=o['MAE']
xxz1=r['MAE']
xxz2=dsx['MAE']
xxz3=p['MAE']

xxxz=o['R2']
xxxz1=r['R2']
xxxz2=dsx['R2']
xxxz3=p['R2']


tab2 = [
    ["Decision Tree",xxz,xxxz,xz],
    ["Random Forest Regressor",xxz1,xxxz1,xz1],
    ["XGBoost",xxz2,xxxz2,xz2],
    ["Linear Regression",xxz3,xxxz3,xz3]
]
 
head = ["Model","MAE","R2","RMSE"]
print(tabulate(tab2, headers=head, tablefmt="grid"))